The Patterns Of Coincidence

Posted by on Apr 27, 2005 | No Comments

This posting might not start like it has anything to do with helping seniors achieve computer literacy, but stick with it for a few paragraphs.

When strange coincidences occur, there might be the making of a column. A long time ago in a previous edition of this series, I mentioned the Norton anti-virus products with specific emphasis on fixing a problem with LU1803, and after a hiatus of several months, I’ve started to get e-mail asking for more details on how I fixed my wife’s machine when she developed the LU1803 problem. For those of you to whom that is a nonsense alphanumeric string of characters, this is a problem with Norton’s LiveUpdate feature which can cause severe heartburn if not treated. The correct treatment is not always obvious.

Why do these things seem to go in cycles? There is no apparent reason why machines that have been running happily would suddenly start getting LU1803 indications, particularly when these computers are in various parts of the world and it’s very likely the operators don’t know each other.

The same thing happens with other topics such as desktop clutter. I will be wondering what to write about next week and suddenly after several months of not hearing any feedback about desktops, I will get 3-5 letters in one day asking about things like the iconoid application. The same thing has happened with questions about mouse control by seniors with arthritis, and whether I can recommend suitable alternative pointing devices.

I can think of only two possible explanations:

1. I’m making it up – that is, our minds are organized to sense patterns in data even when the data is random. That’s why we have readily identifiable constellations in the sky. Everyone can point out the Big Dipper, but the component stars are not actually a close-knit group. (Although those of you blessed with excellent vision might be able to resolve what appears to be one star in the handle into what it is in reality, a double star). So when I need a topic to write about that will be of general interest, the automatic pattern recognition system goes to work scanning my inbox until a pattern emerges, even if there is no pattern in reality.

2. The cycles are real and are produced or amplified by the articles themselves. That is, an article about LU1803 is posted and people with an interest respond over the next few days. That essentially empties the well. Then in the normal course of events, people continue to have LU1803 issues, and it takes a while for the traffic to build up again. At which time I post another column on it, etc.

Similar things happen in teaching a class and even when giving private lessons. The students will seem to be making satisfactory progress on item C after having mastered A and B. Then suddenly you notice that while working on C, several of the class are making silly mistakes utilizing A. That’s odd, they should have A down cold. An unwary instructor might inadvertently embarrass a student by pointing out the error and making light of it. Joking is okay, but a savvy instructor knows that care must be taken to avoid bruising egos. Putting oneself in the hands of an instructor is more difficult for many seniors than it is for younger people who are already immersed in a teaching environment.

Instead of looking for the reason in the students, if a pattern has truly arisen showing that mastery of A is not complete for at least some students, the instructor has to think of a clever way of working it back into the exercises without indicating this is remedial work. That is easier to accomplish than to write about. Almost all exercises done in a typical program will depend on everything that has been learned before. Therefore when learning C, A and B are actually being used, and a slight change in emphasis with a few “As you remember from the class on B…” comments thrown in, the problem can be addressed with minimal disruption.

Another tool that can have similar results is to have regular review sessions. These can be a few minutes at the end of a class, but the key is that they become an expected regular feature of the class. Therefore, nothing is strange if the review dwells on A for longer than usual.

If anyone comments on this posting, I will respond after a measured time depending on the content of the letter. Not that I am trying to break up cycles – why would I do that?

For more in-depth tips on tutoring seniors, see the complete tutorial here. I also have posted a tutorial on elementary decision theory for those who might question a physician’s diagnosis (important for seniors) or anti-terrorist activities (important for everyone) but haven’t had the framework to analyze the data. That tutorial can be found here.